Method and systems for monetizing financial brokerage accounts are disclosed. One aspect for certain embodiments includes mining data from financial brokerage accounts and monetizing the mined data and providing to the customer an unlimited number of free trades for an unlimited period of time.
|
1. A computer-implemented method for monetizing financial brokerage accounts, the method comprising:
storing in an information repository input data received online for a customer of a financial brokerage system;
creating an individual profile for the customer describing attributes of the customer based on the input data of the customer stored in the information repository;
automatically classifying the customer in one or more taxonomies based on the individual profile for the customer, each one of the one or more taxonomies being subject to a distinct set of rules satisfying criteria of the financial brokerage system;
receiving, for each of a plurality of users including the customer, respective first transactional data indicative of one or more transactions involving the respective user and respective first behavioral data indicative of one or more behaviors of the respective user;
executing a clustering algorithm using the first transactional data and the first behavioral data of the plurality of users to generate one or more profile clusters;
mining data from the individual profile for the customer and the one or more taxonomies;
executing a classification algorithm using the mined data in real-time to classify the customer in a first profile cluster of the one or more profile clusters, wherein the first profile cluster monetizes the mined data in real-time to predict customer behavior subject to the distinct set of rules;
providing, via a website over the internet, the customer with free trades on a display screen of a device, wherein the at least one computer module monetizes the mined data by performing data analytics including automatically generating in real-time a new set of rules to learn customer behavior and to predict subsequent customer behavior, the new set of rules being generated in response to change in attributes of the customer;
receiving, for the user, at least one of second transactional data or second behavioral data subsequent to predicting the customer behavior;
updating in real-time over the internet on an on-going basis at least one of the clustering algorithms or the classification algorithm based on the predicted customer behavior and the at least one of the second transactional data or the second behavioral data, wherein the updating includes modifying the set of rules for the classification algorithm;
and re-classifying, based on the at least one of the updated clustering algorithms or the updated classification algorithm, the customer in a second profile cluster.
2. The computer-implemented method of
3. The computer-implemented method of
4. The computer-implemented method of
5. The computer-implemented method of
6. The computer-implemented method of
7. The computer-implemented method of
8. The computer-implemented method of
9. The computer-implemented method of
10. The computer-implemented method of
11. The computer-implemented method of
12. The computer-implemented method of
13. The computer-implemented method of
14. The computer-implemented method of
15. The computer-implemented method of
16. The computer-implemented method of
17. The computer-implemented method of
18. The computer-implemented method of
19. The computer-implemented method of
automatically updating in real-time the classification of the customer from a first taxonomy of the one or more taxonomies to a different second taxonomy of the one or more taxonomies based on change in the attributes of the customer; and
predicting the second taxonomy based on the distinct set of rules.
20. The computer-implemented method of
automatically integrating the new set of rules into the distinct set of rules based on validation of the new set of rules over a period of time.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
This present application is a continuation of U.S. patent application Ser. No. 16/232,731 filed Dec. 26, 2018, which is a continuation of U.S. patent application Ser. No. 14/589,302 filed Jan. 5, 2015, which is a continuation of U.S. patent application Ser. No. 13/489,301 filed Jun. 5, 2012, which claims priority to U.S. Provisional Application No. 61/636,508 filed Apr. 20, 2012. All applications are incorporated by reference in their entirety herewith.
The disclosed embodiments relate generally to data analytics. More particularly, the disclosed embodiments relate to monetizing financial brokerage data.
For a better understanding of the aforementioned aspects of the invention as well as additional aspects and embodiments thereof, reference should be made to the Description of Embodiments below, in conjunction with the following drawings in which like reference numerals refer to corresponding parts throughout the figures.
Methods, systems, user interfaces, and other aspects of the invention are described. Reference will be made to certain embodiments of the invention, examples of which are illustrated in the accompanying drawings. While the invention will be described in conjunction with the embodiments, it will be understood that it is not intended to limit the invention to these particular embodiments alone. On the contrary, the invention is intended to cover alternatives, modifications and equivalents that are within the spirit and scope of the invention. The specification and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense.
Moreover, in the following description, numerous specific details are set forth to provide a thorough understanding of the present invention. However, it will be apparent to one of ordinary skill in the art that the invention may be practiced without these particular details. In other instances, methods, procedures, components, and networks that are well known to those of ordinary skill in the art are not described in detail to avoid obscuring aspects of the present invention.
According to certain embodiments, detailed personal information and financial information of a user of a financial brokerage account is cross referenced with the user's investment patterns, investment preferences and other online behavior of the user for purposes of data analysis and monetization.
According to certain embodiments, analytics of the user's financial brokerage account information and brokerage transactional behavior can be profiled to correlate dynamically with current world events to trigger investment offers and or commercial sales offers and or service offers to the user.
According to certain embodiments, analytics of a given user's financial brokerage account information and brokerage transactional behavior can be profiled for searches for additional data from public records on the given user or on other users with profiles similar to the given user to provide for further data analytics on the given user to trigger investment offers and or commercial sales offers and or service offers to the user.
According to certain embodiments, data analytics are used for building profiles for individual financial brokerage account holders as well as building profiles that are based on brokerage data and or other third party data that meet certain criteria for purposes of monetization in various industry segments.
According to certain embodiments, value may be returned to users of the financial brokerage account through free trades, money back, credits, coupons, promotional or other rewards programs. Free trades can also be provided to referral clients referred by a customer (user) of the financial brokerage account.
According to certain embodiments, a user need not fund his/her account in the financial brokerage. In other words, a user can open a “non-funded” brokerage account. The user can merely provide his/her detailed personal information and financial information to the financial brokerage. Such information can be used for data analysis and monetization.
According to certain embodiments, data analytics of the user data include building various data taxonomies, user classification profiles, profile clusters, and rules engine for predicting behavior or that satisfies criteria for various market segments.
According to certain embodiments, free trades include commission-free trades. The customer is not charged a sales charge or commission associated with the customer's brokerage transaction.
According to certain embodiments, free trades include spread-free trades. In such trades, the spread between the bid and ask price is eliminated.
According to certain embodiments, free trades include one or more of the following: 1) no load on mutual fund transactions, 2) eliminating principal's fees on fixed income securities transactions, and 3) eliminating principal's fees on equity securities transactions.
According to certain embodiments, value can be returned to the user/customer of the financial brokerage account by providing the user/customer unlimited number of free trades. Also, value can be returned to the user/customer of the financial brokerage account by providing the user/customer unlimited number of free trades for an unlimited period of time.
According to certain embodiments, in addition to providing free trades, value may be returned to users/customers of the financial brokerage account by providing one or more of the following: 1) free check writing, 2) free bill payments, 3) eliminating annual fees on the customer's account, 3) eliminating transfer fees for the customer's account, 4) eliminating fees for debit card transactions, 5) eliminating fees for credit card cash advances, 6) eliminating fees on the customer's retirement account, 7) eliminating maintenance fees for employee stock option plans, 8) eliminating wire transfer fees for the customer's account, 9) eliminating ATM fees for the customer's account, and 10) eliminating ACH fees for the customer.
According to certain embodiments, in addition to providing free trades, value may be returned to users/customers of the financial brokerage account by eliminating fees for market data services. Market data services can include providing real time quotes and/or streaming quotes to the customer/user of the financial brokerage account.
According to certain embodiments, value can be returned to a user who has an account (whether funded or unfunded) in the financial brokerage by providing an online shopping site associated with the financial brokerage and where the brokerage user can shop and accumulate credits. Further, data analysis can be performed on the user's shopping behavior in conjunction with the user's personal and financial information. Users can shop by redeeming credits/coupons provided by the financial brokerage as a return of value to the user.
According to certain embodiments, a user/customer of the financial brokerage account may be any one of the following non-limiting entities: an individual, a for-profit business entity, a charitable organization, a scholastic institution, a public works entity, a government entity, a sovereign wealth fund, a trust fund, or an institutional investor.
New Account data: According to certain embodiments, some non-limiting examples of new account data include the following.
Margin Account data: According to certain embodiments, some non-limiting examples of margin account data include the following.
Options Account data: According to certain embodiments, some non-limiting examples of options account data include the following.
Retirement Account data: According to certain embodiments, some non-limiting examples of retirement account data include the following.
ACH data: According to certain embodiments, some non-limiting examples of ACH account data include the following:
Investment Access data: According to certain embodiments, some non-limiting examples of investment access data include the following.
Certification of Investment Power (Corp) data: According to certain embodiments, some non-limiting examples of certification of investment power (Corp) data include the following.
Certification of Investment Power (Trust) Data: According to certain embodiments, some non-limiting examples of certification of investment power (Trust) data include the following.
RIA data: According to certain embodiments, some non-limiting examples of RIA data include the following.
$105,000
83%
17%
$106,200
79%
21%
$107,400
76%
24%
$108,500
73%
27%
$109,500
71%
29%
Tax data: According to certain embodiments, some non-limiting examples of tax data include the following.
W-9
Transactional data: According to certain embodiments, some non-limiting examples of transactional data include the following.
Traded on the following market centers
Behavioral data: According to certain embodiments, some non-limiting examples of behavioral data include the following.
Credit check and address change: In addition, the brokerage system collects and stores credit check results and address change information associated with the customers/users of the brokerage system for purposes of data analytics and monetization, according to certain embodiments.
Various data elements are organized before applying techniques of data analytics. A financial brokerage system has many types of information about the customer/user of the financial brokerage system. Some of the information is gathered at the time of registration and others accumulated over time based on the types of transactions the customer performs at the web site. For example, all the activity of the customer on the online financial brokerage system over time including time and frequency of visits to the online financial brokerage system will be captured by the financial brokerage system. In addition, the financial brokerage system can take advantage of data from outside sources to learn more about customer expectations and behavior. This may require working out the right agreements with partners for promotions and advertisements. For example, such data can include the purchase behavior of certain items like cars or insurance from ad networks or car dealers, for a profile of users that are potential buyers of such items that the financial brokerage system can identify with its own data. Other examples of data sources can include news feeds, court electronic databases (PACER), Department of Motor Vehicles, Internet, community multimedia centers, social networks, etc. In other words, the data from data sources outside of the financial brokerage system can be used in conjunction with brokerage data for purposes of data analytics. The data organization and analytics techniques and monetization objectives of the financial brokerage system can inform on the types of combinations of outside data sources with data from the financial brokerage system for monetization. The following non-limiting examples illustrate some useful data analytics of brokerage data in combination with other sources of data for predicting behavior or for making suggestions or offers to the customer of the financial brokerage account or for sharing information with various industry partners for purposes of monetization of data. The following examples are merely illustrative and are for purposes of explaining the concept of combining various sources of data with brokerage information in order to generate revenue for the financial brokerage company and to return value to the customer.
UNESCO and Al Jazeera to promote freedom of expression in the Arab World.+marital status of customer (married)+joint account+H income of customer (high)+of customer net worth (high)+W8 (non US Citizen from Egypt)+address of customer (indicating US main address)+4 dependents of customer+direct deposits into 3 custodial savings accounts of customer+address of customer=a donation to United Nations Educational Scientific and Cultural Organization.
The Pope visits Cuba, greeted by Raul Castro and later meets with Fidel Castro.+zip code (Miami Florida high latino density area)+income of customer+net worth of customer+wire transfer requests (to The Cuban American National Foundation)+account log in of customer and trading records of customer+occupation of customer (owns religious book store or professor at Christian school)=donation to Church (donation size varies based on income and net worth fields).
News example: batteries in electric cars are catching on fire+customer came to our site after visiting a hybrid car web site=change to purchase Hybrid car for customer
Justin Bieber is playing in LA+high net worth of customer+high income of customer+zip code+Occupation of customer (attorney for Columbia Records)+2 custodial accounts of customer+age of custodian (11 and 13 years of age)+birthday of custodian around time of concert=purchase Justin Bieber tickets.
An anti war pro social programs president+income of customer+net worth of customer+zip code of customer+occupation of customer+tax bracket of customer+customer has defense stocks in portfolio=reduce and or sell military stocks in your portfolio.
Monitor Public Access to Court Electronic Records (PACER) www.pacer.gov “PACER is an electronic public access service that allows users to obtain case and docket information from federal appellate, district and bankruptcy courts and the PACER Case Locator via the internet.
Court ruling deciding in favor of a challenge to a recent board decision in some town USA to only allow 1 house per 5 acre zoning instead of what had always been 1 house per 2 acre zoning.+gender of customer (male)+marital status of customer (married)+income of customer+net worth of customer+3 dependents+zip code of customer+occupation of customer (owns residential construction company)+college savings account of customer+also has business account (residential builders)+applied for special purpose business loan to build a 50 unit residential complex (we know this because he either applied for the loan through us or the bank he is getting the loan from is contacting us to verify the assets in his account held with us)+interest rates reported by Yahoo finance to be low for years=building 100 residential units instead of 50.
3 speeding tickets of customer, one of which was reckless driving and a suspended license+income of customer+net worth of customer+date of birth of customer (birthday in 1 week)+zip code of customer+investment objectives of customer (preservation of capital shows the person is conservative)+marital status of customer+bought tickets online from our points program for Scotch tasting event on birthday or visited Scotch taste testing site prior to or after coming to our website=needs a car and driver.
Recent DWI of customer+marital status of customer (married)+income of customer+net worth of customer+current date (New Years Eve)+zip code of customer+3 dependents=rent limousine.
Monitor FINRA broker check—financial industry regulatory authority. FINRA BrokerCheck—is a free online tool to help investors check the professional
Customer's investment broker (who has multiple infractions including unauthorized trades)+marital status of customer+income of customer+net worth of customer+zip code of customer+investment objectives of customer+address of customer+current date+(all fields filled out)=transfer account to new financial brokerage firm
News of a severe storm headed in customer's direction (area New Orleans)+marital status of customer+income of customer (low)+net worth of customer (low)+customer owns home+2 dependents+zip code of customer (lower 9th ward)=purchase of ply wood to board up windows.
News of a severe storm headed in customer's direction (area New Orleans)+marital status of customer+income of customer (high)+net worth of customer (high)+customer owns home+2 dependents+zip code of customer (affluent suburb of New Orleans)=suggest purchase upscale hotel room for two weeks.
News of a severe storm headed in customer's direction (area New Orleans)+marital status of customer (single)+income of customer (unemployed)+net worth of customer (high)+customer rents+zip code (apartment on Bourbon street)=take two week vacation in Jamaica (Jamaica because of previous intelligence gathered by financial brokerage system)
Public records show, customer paid 100 k for house+credit report (states balance of loan 85 k)+indicates down payment below 20% on purchase=mandatory PMI (private mortgage insurance)+zillow appraisal (value of house)+lower interest rates+value of all accounts held by customer+zip code of customer+income of customer+net worth of customer+conservative investment objectives of customer+refinancing=New refinance (valuable to mortgage brokers also because of knowledge that mortgage brokers can now save the customer additional money by eliminating the PMI)
An important feature of monetizing customer behavior is the creation of one or more profiles of the customer of the financial brokerage system. The profile describes the customer's attributes. The profile can be used to provide the right advertisement or promotion when the customer is logged in. The profile can also be used to provide a list of customers to ad networks or other partners who are interested in offering some special promotions. New profiles can be created easily based on partner requirements. For example, a car dealer may have certain criteria that the profiles need to satisfy. For this level of flexibility, one or more taxonomies are created to classify the financial brokerage system customers for different profile needs.
The individual profile can be provided to marketers at different granularities to target advertisements. Profile clusters will be useful to marketers for making special promotions or offers. The financial brokerage system can give marketers the list of customers that fit into specific profile clusters with a measure of confidence (for example, 1-10). Further, marketers can work the financial brokerage system to create special profiles. The taxonomy and individual profiles will facilitate quick and easy way of creating new profile clusters.
Another feature of data analytics is around clusters of profiles. Profile clusters identify groups of customers/users that fit some specific profiles of interest.
Further, rules are generated to apply the individual profiles, attributes in taxonomy or profile clusters to test and learn customer behavior and thus predict future behavior. The rules can be used in multiple ways:
Methods of data mining include but are not limited to customer/user classification taxonomy, individual customer/user profile generation, profile cluster development and rules generation, regression analysis and learning & prediction.
Customer Classification Taxonomy: Create taxonomy that helps classify customers for different characteristics. A customer can belong to multiple nodes in the taxonomy. The taxonomy can be used to create individual profiles for quick access and to generate lists of customers that meet specific characteristics. The taxonomy is created after reviewing the different customer data items in the financial brokerage system. The following is a non-limiting example of a taxonomy. The taxonomies are based on the objectives of the financial brokerage system and can vary from implementation to implementation.
Gender
Male
Female
Age
< 30 years
30 - 40
40 - 50
50- 60
60 -70
> 70
Marital Status
.....
.......
Annual income
< $100K
.......
Net worth
< $500K
$500K - $1M
.....
Occupation
......
Zip Code
Average household income:
< $500K
$500K - $1M
....
Or Average house value:....
......
Investment Objective
Preservation of Principal
Conservative Growth
Aggressive
Speculation
......
Risk Profile
.....
Spouse’s Occupation
.....
Spouse’s Age
.....
Country of citizenship
.....
Tax bracket
.......
Employer Zip Code
.......
Stocks
.....
Trading style
Daily
Weekly
Monthly
.....
Children
Elementary school
1
2
3
> 3
Middle school
1
2
....
High school
...
College
....
....
Individual Customer Profile: For each customer all available information are stored (both internally available from the customer's use of the financial brokerage system and gathered from outside sources) in the profile. The data associated with the customer classification taxonomy are stored in the profile in the format that matches the nodes in the taxonomy. This taxonomy related information will be stored to match profile queries in an efficient manner.
Profile Clusters: Clusters are special profiles that match certain characteristics of interest to marketers, for example, based on industry behavior. Each cluster can have its own identifier or name. Each customer can be classified under multiple clusters. The following characteristics are merely examples for purposes of explanation.
Rules: Data analytics includes defining rules that predict behavior based on individual customer profile or profile clusters. The system can trigger alerts or action when the rules are satisfied. The following rules are merely examples for purposes of illustration.
Customers can be classified for different partner needs or behavior prediction by learning through rules and clustering. Partners of the financial brokerage system can include commercial retailers, banks, advertising agencies, insurance companies, realtors, mortgage brokers, airline industry, entertainment industry, hospitality industry, recreation industry, etc. Both associative rules and decision trees will be used. Regression also will be used for learning and prediction.
The techniques described herein (rules and decision trees for classifications, clustering and regression) can be used for predictive analytics. Such techniques can be used when new accounts are created and when attributes for existing accounts change. The expected behavior of a new customer that fits into certain profile (e.g., a single parent with small children who just changed address will be expected to look for day care services) can be predicted using the rules that have been created in the system. Similarly when an attribute of the customer changes, the customer may fall into a new cluster and a new set of rules will apply to predict certain behavior. For example, for a customer that crosses a certain age (e.g., 65 years of age), new offers can be provided to this customer based on the rules in place for the new cluster.
Rules Learning: Rules are used in learning systems to predict behavior. The rules are looked up for the right conditions to decide the outcome. The antecedent or precondition of a rule is a series of tests, while the consequence or conclusion give the class or classes that apply to instances covered by the rule, or sometimes a probability distribution over the classes. The preconditions are logically ANDed together, and all the tests must succeed in order for the rule to trigger.
Examples of decision tree: The following is merely an example for purposes of illustration.
Net worth −> Age range −> Number of children => expected education
expense
Has power of attorney −> age of account holder => need for assisted
living
Married Status
Single
Income ( > $x)
Number of dependents (> 0)
Custodian Account (Yes)
Age of children (< 12)
Use day care center
Married
Income ( > $Y )
Number of dependents ( >0 )
Custodian Account ( Yes )
Age of children ( 5 -18 )
Use private school
Association rules: The following association rules are examples for purposes of explanation.
Initial set of rules are created manually based on expertise and hypotheses. These rules can be created as decision trees or tables. 75% of existing information can be used for preparing these rules. This data is the training data set. The remaining 25% of the data is used for testing the rules. For example, consider the following sample rules:
Single+income+net worth+change of address+number of dependents+custodian account+age of minor=need for new day care center.
Married+high income+high net worth+change of address+number of dependents+custodian account+age of minor=need new private school
The data values at each decision point for the above are created using the training data set and validated using the remaining data. The correlation results can be measured and can be visually presented to verify the correctness of the rules.
Clustering: Clustering is the process of discovering groups and structures in the data that are in some way “similar”, without using known structures in the data.
Hierarchical clusters using information about gender, profession, net worth, number of children, stocks, etc., can be created. Probability based clustering can be used. Clustering can be directed with different types of seed data to evaluate the effectiveness of the resulting clusters. Seed data are initial sets of data to act as centroids of the different clusters. As a non-limiting example, if clusters are to be directed by net worth followed by investment risk profile, then clustering can be initiated with chosen sets for net worth and possibly randomly chosen investment risk profiles. Through multiple iterations, clustering will stabilize around centroids.
Classification: Classification is the process of identifying groups or rules to apply for new data items based on rules already created using examples. The financial brokerage system can use both association rules and decision trees for classification. For decision trees, one rule is generated for each leaf. The rules will be unambiguous in that the order of execution of the rules is irrelevant. To classify customers into different classes, association rules can be used.
When new customer data is entered into the financial brokerage system, the classification rules will be applied to identify the classes to which the customer belongs. For example, by using the decision trees, the customer can be classified as one needing day care center, need for assisted living, etc. Similarly, the association rules can be used to classify customers as high or low risk investors, those who buy things for children, etc.
Regression: Regression can be used for numeric data that fit statistical functions. Most commonly used techniques involve linear regression where the right fitting function can be derived using statistical packages. The idea is to express the class as a linear combination of attributes, with pre-determined weights. For example, one situation where regression can be applied is to find a pattern between the number of transactions by the financial brokerage customers and login time before and after market close by the customers. If there is a relationship, one can predict the number of transactions for those customers logging in to the financial brokerage system at certain periods of time.
Some non-limiting examples of using data from external sources for predicting potential behavior and thus monetize the customer profiles already created are given below:
It is important to present the results of data analytics in a way that is easy to understand and that summarizes the results for the target audience. Different reports and charts and graphs can be prepared using standard reporting packages. Some sample reports can include the number of new customers that fit within the different classifications or profiles and the trends over a period of time. Another report can be the success rate of the different rules for classifications.
Heat map is another effective visual presentation to show data relationships in multiple dimensions. A heat map presents data within classes (or matrix) as colors. The size of each matrix represents the relative importance one attribute and the color can represent the relative importance of another attribute. For example, a representation of clusters around net worth in a heat map can represent the size of each cluster by the size of the rectangle and the color can indicate the number of transactions performed by each cluster. Red can indicate high activity and yellow low activity. Visually, such a representation is easy to understand and can easily identify which clusters have the most activity.
After the initial data models (classification taxonomy, rules, clusters and regression functions) are created, the system needs to be maintained on an on-going basis.
The initial taxonomy can be created manually by reviewing the customer data by the financial brokerage system internal experts. This can be enhanced over a period of time as the system learns more about customer behavior patterns and if higher degrees of granularity of profiles are needed.
Once the taxonomy is created individual profiles can be derived from it. An automated process can maintain the profiles on a regular basis. This will be a matter of looking up data values for each node in the taxonomy and keeping a bit pattern for the profile.
Clusters are created in different ways. Using the data of known industry behavior of interest of customers (those buying high risk instruments, those who put on a hedged position, etc.) a set of industry behavior clusters will be created.
Another set can be clusters of interest for predicting industry behavior of known groups or test hypotheses (those who log in into account 30 minutes prior to market close and 30 minutes after market close likely to perform transaction, those who check their account every day are likely to put on a hedged position, those who buy only funds likely to stay only within those bonds, etc.).
Clusters for non-industry behavior are created using hypotheses and on marketer's (or business partner of brokerage) request. Non-limiting examples can include:
Many of the clusters can be created automatically using the classification rules and association rules. In addition K-means clustering can be used to create new clusters using cluster centers chosen based on industry experience (as a non-limiting example, net worth+age). The data set can be split for building the models as well for testing them. Results of special promotions or targeted advertisements at these clusters will validate the hypothesis as well as refine the clusters.
Initial set of rules are created manually. Profile clusters can be used for generating rules. Some of the rules will be used in conjunction with clusters to trigger events. An example will be change in address for those in single parent with small children clusters needing new day care. The system will alert this change which will be used for sending special offer or promotion to the customer.
Creation of the rules and clusters is an iterative process. The data set aside outside of the training data set can be initially used for validations. For different rules and clusters, the expected precision and recall values can be defined. Tests will show how well the models work for the required precision and recall values. The models can be tuned to fit the expectation. The system can regularly monitor the effectiveness of the prediction against the rules using the precision and recall values as the guide.
New association rules can be identified as part of the learning process (e.g., a pattern about the timing of the change of address of single parent family can predict such changes in the future and hence certain behavior for purchases. As another example, some number of members of a cluster of ‘risk investors’ tend to purchase some types of instruments at certain periods of time, etc.). Some rules that include data from external sources may be difficult to validate, initially. But over a period of time enough data would be collected to validate the rules. Data provided by business partners of the financial brokerage system can be used to improve the rules in such cases.
Profiles can be automatically generated for new customers and appropriate clusters can be identified. Promotions or special offers can be made to these new customers. The new customers' behavior can be tracked against the rules to increase learning and improve the prediction in the system.
Data that does not fall into any category through any of the methods like classification, association, regression and clustering can be of significance. Such data can be reviewed to see if the data is really an anomaly or better tuning of the classification models can help identify new classes to fit such data. The precision and recall values in classification and clustering can be adjusted to experiment. New refined clusters can be created or rules can be modified for classification. This can be part of the on-going maintenance of the system.
The foregoing description, for purpose of explanation, has been described with reference to specific embodiments. However, the illustrative discussions above are not intended to be exhaustive or to limit the invention to the precise forms disclosed. Many modifications and variations are possible in view of the above teachings. The embodiments were chosen and described in order to best explain the principles of the invention and its practical applications, to thereby enable others skilled in the art to best utilize the invention and various embodiments with various modifications as are suited to the particular use contemplated.
| Patent | Priority | Assignee | Title |
| Patent | Priority | Assignee | Title |
| 8433645, | Aug 04 2010 | PORTWARE, LLC | Methods and systems related to securities trading |
| 20100325062, | |||
| 20120221486, |
| Executed on | Assignor | Assignee | Conveyance | Frame | Reel | Doc |
| Date | Maintenance Fee Events |
| Jun 10 2022 | BIG: Entity status set to Undiscounted (note the period is included in the code). |
| Jun 16 2022 | SMAL: Entity status set to Small. |
| Date | Maintenance Schedule |
| Aug 29 2026 | 4 years fee payment window open |
| Mar 01 2027 | 6 months grace period start (w surcharge) |
| Aug 29 2027 | patent expiry (for year 4) |
| Aug 29 2029 | 2 years to revive unintentionally abandoned end. (for year 4) |
| Aug 29 2030 | 8 years fee payment window open |
| Mar 01 2031 | 6 months grace period start (w surcharge) |
| Aug 29 2031 | patent expiry (for year 8) |
| Aug 29 2033 | 2 years to revive unintentionally abandoned end. (for year 8) |
| Aug 29 2034 | 12 years fee payment window open |
| Mar 01 2035 | 6 months grace period start (w surcharge) |
| Aug 29 2035 | patent expiry (for year 12) |
| Aug 29 2037 | 2 years to revive unintentionally abandoned end. (for year 12) |